Striking Out? The Economic Impact of Major League Baseball WorkStoppages on Host Communities
Victor A. Mathesonand
Robert A. Baade
April 2005
COLLEGE OF THE HOLY CROSS, DEPARTMENT OF ECONOMICSFACULTY RESEARCH SERIES, WORKING PAPER NO. 05-07*
Department of EconomicsCollege of the Holy Cross
Box 45AWorcester, Massachusetts 01610
(508) 793-3362 (phone)(508) 793-3710 (fax)
http://www.holycross.edu/departments/economics/website
*All papers in the Holy Cross Working Paper Series should be considered draft versions subjectto future revision. Comments and suggestions are welcome.
†Victor A. Matheson, Department of Economics, Box 157A, College of the Holy Cross,Worcester, MA 01610-2395, 508-793-2649 (phone), 508-793-3710 (fax),[email protected]
††Robert A. Baade, Department of Economics and Business, Lake Forest College, LakeForest, IL 60045, 847-735-5136 (phone), 847-735-6193 (fax), [email protected]
Striking Out? The Economic Impact of Major League Baseball WorkStoppages on Host Communities
Victor A. Matheson†
College of the Holy Cross
and
Robert A. Baade††
Lake Forest College
April 2005
Abstract
Major League Baseball teams have used the lure of economic riches as an incentive forcities to construct new stadiums at considerable public expense. Estimates of the economicimpact of a MLB on host communities have typically been in the vicinity of $300 million. Ouranalysis suggest these numbers are wildly inflated. Using the baseball strikes of 1981, 1994, and1995 as test cases, we find the net economic impact for a MLB team on a host city of $16.2million under one model and $132.3 million under a second model.
JEL Classification Codes: L83
Keywords: impact analysis, baseball, sports economics
1
Introduction
Professional baseball is big business in the United States. Major League Baseball (MLB)
attracts over 70 million fans to games each year with television viewing audiences many times
this number. Top players routinely receive contracts that pay them in excess of $10 million per
year. The construction of Camden Yards in 1992 has prompted a boom in stadium construction
that has seen 14 new stadiums being completed during the 1992-2002 period with new stadiums
being proposed for another 11 cities (Munsey and Suppes, 2000). The majority of the cost of
these construction projects has been bourne by local taxpayers.
Many attempts have been made to quantify the economic impact that the presence of a
MLB team has on a host city. Interest groups wishing to develop public backing for attracting a
new team to a metropolitan area or existing Major League teams who wish to cultivate voter
support for public financing of new stadium facilities frequently commission economic impact
studies that report large contributions from professional sports teams to local economies. For
example, the Oregon Baseball Campaign, a group dedicated to bringing MLB to Portland,
reported that “a MLB team and ballpark would generate between $170 and $300 million
annually in gross expenditures to the state of Oregon.” (Oregon Baseball Campaign, 2002) A
similar study completed for the Virginia Baseball Authority stated that a “a major league
baseball franchise and stadium in northern Virginia would pump more that $8.6 billion into the
economy over 30 years,” or $287 million annually. The St. Louis Regional Chamber and Growth
Association estimated that the Cardinals brought $301 million in economic benefits to the region
with another potential $40 to $48 million in benefits from a post-season appearance. (St. Louis
Regional Chamber and Growth Association, 2000) The lowly Montreal Expos, with attendance
2
less than one-third that of the typical MLB team, produce “a GDP of $105.3 million” according
to a federal report on sport in Canada. (Beaudry, 2002) Even pre-season games apparently
generate significant economic activity. The Florida Sports Foundation announced that the
Grapefruit League, the Spring Training league for 20 of the 30 MLB teams, generated a $490
million impact to the state from 306 games attracting just over 1.5 million fans in 2002. This
combined attendance figure represents an amount roughly two-thirds that which the average
MLB team attracts during the regular season.
Public finance economists, on the other hand, are in general agreement that the figures
produced by sports boosters are wildly inflated. Many studies, including Baade and Dye (1990),
Rosentraub (1994), Baade (1996), Noll and Zimbalist (1997), and Coates and Humphreys (1999)
to name just a few, have examined the economic impact of stadium construction. Without
exception, these studies have found that new stadiums provide little or no net economic stimulus
to the communities in which they are located. Others such as Porter (1999) and Baade and
Matheson (2000) have examined sports “mega-events” such as the Super Bowl and the MLB
All-Star Game. Porter used regression analysis to determine that the economic impact of the
Super Bowl on the host city was statistically insignificant, that is not measurably different from
zero (Porter, 1999). Likewise, Baade and Matheson (2000) challenged an MLB claim that
annual All-Star Game contributes $75 million to the host city economy. Their study of taxable
sales data and employment data concluded that the All-Star Game was actually associated with
lower than expected economic activity for host cities. Again, these researchers find that boosters’
estimates of the economic impact of large sporting events exaggerate the true economic impact
of these events by up to a factor of ten.
3
But the question remains, how big is the true economic impact of professional sports
teams on their host cities? In particular, this paper will examine Major League Baseball teams’
economic contribution to their local metropolitan areas using the 1981 and 1994/95 baseball
strikes as test cases.
Review Sports Economic Impact Studies
The economic numbers quoted by baseball promoters are usually generated using a
standard expenditure approach to estimating the direct economic impact of the event. The
numbers are derived by estimating the number of “visitor days” as a result of the team and
multiplying that statistic by the average estimated per diem expenditures per visitor. Once an
estimate of direct impact is obtained, the total economic impact is estimated by applying a
multiplier which typically doubles the direct economic impact. Using this technique, if a mistake
is made in estimating direct expenditures, those errors are compounded in estimating indirect
expenditures. The secret to generating credible economic impact estimates using the expenditure
approach is to accurately estimate direct expenditures. A annual figure of $300 million for a
MLB team could be roughly arrived at simply by assuming 2.5 million fans per year (only
slightly more than the MLB average) each spending an average of $60 on their visit and then
applying an economic multiplier of 2.
Precisely measuring changes in direct expenditures is fraught with difficulties, however.
Most prominent among them is an assessment of the extent to which spending in conjunction
with the event would have occurred in the absence of it. For example, if an estimate was sought
on the impact of a professional sports team on a local economy, consideration would have to be
4
given to the fact that spending on the team may well merely substitute for spending that would
occur on something else in the local economy in the absence of the event. As pointed out by
Andrew Zimbalist when discussing the 2000 “Subway Series” between the New York Mets and
New York Yankees, “If you buy a $100 ticket to the Series, that’s money you might have spent
on a Broadway show or food.” Therefore, if the fans are primarily indigenous to the community,
a MLB team may simply yield a reallocation of leisure spending while leaving total spending
unchanged. This distinction between gross and net spending has been cited by economists as a
chief reason why professional sports in general do not seem to contribute as much to
metropolitan economies as boosters claim (Baade, 1996). There is nearly universal agreement
among independent economists that spending by local residents must be excluded from
economic impact calculations due to this substitution effect.
For example, only $37.9 of the reported $105.3 million estimated impact of the Montreal
Expos comes from outside of Montreal. Similarly, the St. Louis Cardinals report that only 32%
of their fan base comes from outside the St. Louis metropolitan area and that percentage is one of
the highest in the league. (St. Louis Regional Chamber and Growth Association, 2000) The
large economic impact figure for the Grapefruit League is largely a result of the fact that
spending on this league is more likely to be categorized as export spending since most of it is
thought to be undertaken by people from outside the communities with 60% of spring training
fans being visitors to the Florida region.
At first blush, excluding local spending from economic impact analyses should eliminate
the upward bias in the calculations. An examination of mega-events, which tend to have an even
larger percentage of visitors coming from outside the host region (87% of attendees at the 1999
5
Super Bowl were from outside the host city of Miami), still shows that booster claims are far
above ex post analyses of the actual economic effects of these events.
Spending by local residents, therefore, is not the only potentially significant source of
bias in estimating direct expenditures. While surveys on expenditures by those attending a
sporting event complete with a question on place of residence, may well provide insight on
spending behavior for those patronizing the event, such a technique offers no data on changes in
spending by residents not attending the event. It is conceivable that some residents may
dramatically change their spending during an event in order to avoid the congestion in the
venue’s environs. Similarly, while hotel rooms during a local team’s home stand may be filled
with baseball fans, if hotels in the host city are normally at or near capacity during the time
period in which the team is playing at home, it may be that sporting event visitors are simply
crowding out other potential visitors. In general, a fundamental shortcoming of economic impact
studies is not with information on spending for those who are included in a direct expenditure
survey, but rather with the lack of information on the spending behavior for those who are not.
Even out of town visitors who attend a sporting event may not improve the local
economy if their attendance at the game displaces other activities in the city that the visitors
would have done instead. For example, each year at the Western Economic Association
meetings, the conference organizes a baseball outing for attendees. While these visitors to the
conference city undoubtably spend money at the baseball game, these same visitors would have
spent money going out to dinner or to another cultural attraction in the absence of the baseball
game. In other words, even though the sports franchise induces visitor spending, it does not
induce any new spending in the city. In this case, the baseball team does not add economic
6
activity to the city but simply reallocates spending from one area to another.
A second potentially significant source of bias in economic impact studies relates to
leakages from the circular flow of spending. For example, if the host economy is at or very near
full employment or if the work requires specialized skills, it may be that the labor essential to
conducting the event resides in other communities. To the extent that this is true, then the
indirect spending that constitutes the multiplier effect must be adjusted to reflect this leakage of
income and subsequent spending.
Labor is not the only factor of production that may repatriate income. Even if hotels
experience higher than normal occupancy rates during a sporting events, then the question must
be raised about the fraction of increased earnings that remain in the community if the hotel is a
nationally owned chain. In short, to assess the impact of mega-events, a balance of payments
approach must be utilized. Since the input-output models used in even the most sophisticated ex
ante analyses are based on fixed relationships between inputs and outputs, such models do not
account for the expenditure complications associated with full employment and capital
ownership noted here.
As an alternative to estimating the change in expenditures and associated changes in
economic activity, those who provide goods and services directly in accommodating the event
could be asked how their activity has been altered by the event. Unfortunately, most business
managers are unable to accurately predict how much economic activity would have taken place
without the event.
Since the expenditure approach to projecting the economic impact of mega-events is
most commonly used by league and city officials to generate economic impact estimates, we will
7
be comparing the results generated by our model to the estimates quoted by league officials that
were derived using an expenditure approach. In the next sections of the paper, the models that
are used to estimate the impact of the MLB are detailed.
Model #1
The economic activity generated by a MLB team is likely to be small relative to the
overall economy, and isolating the team’s impact, therefore, is not a trivial task. The largest
economic estimates of $300 million represent only 0.6% of the personal income of even the
smallest MLB markets (Fort Worth, Kansas City, Milwaukee, and Cincinnati) and only about
0.1% of the largest MLB markets (New York City, Los Angeles, and Chicago).
An additional difficulty posed by MLB is the fact that its anti-trust exemption has
prevented significant movement in teams. The last team to relocate was the Texas Rangers who
moved from Washington, D.C. in 1971. This provides few opportunities to examine the
economic impact of the loss of a team. Several teams have been added through expansion in the
past 30 years: the Seattle Mariners and Toronto Blue Bays in 1977, the Colorado Rockies and
Florida (Miami) Marlins in 1993, and the Tampa Bay Devil Rays and Arizona Diamondbacks in
1998. While it is certainly possible to examine these cities to attempt to estimate the effect of
adding a MLB team, one may run into an endogenous variable problem in the analysis.
Certainly MLB bases its decision on which cities to grant expansion franchises at least in part on
the economic growth prospects of the applicant cities. Therefore, the question arises: if a city
experiences rapid economic growth following being granted an expansion franchise, is the
growth the result of the franchise or was the franchise granted because of good prospects for
8
economic growth in the city?
With these problems inherent in estimating the economic effects of a MLB franchise in a
metropolitan area, it is natural to search for other ways to measure these effects. The 1981 and
1994 MLB strikes provide natural experiments. The 1994 players’ strike resulted in the
cancellation of 669 regular season games (29.5% of the total season) as well as the entire post-
season including the World Series. The strike also resulted in the loss of 252 regular season
games in the 1995 season (9 of 81 normally scheduled home games per team). In addition, the
ill-will prevailing among baseball fans that resulted from the 1994 strike and the cancellation of
previous year’s World Series caused a 29% reduction in overall baseball attendance in 1995
compared with the non-strike year of 1993. The 1981 strike resulted in the loss of 717 regular
season games (34% of the total), but the strike was resolved before the post-season so these
games were not lost in 1981. Perhaps due to this fact, MLB suffered no let down in attendance
in the following year unlike after the 1994 strike. Eighty-six regular season games were also lost
due to a work stoppage in the 1972 season. The total number of games lost in 1972 represents
less than 5% of the scheduled games and attendance dropped by less than 10% compared to the
preceding season, so this episode will not be considered in this paper.
It is reasonable to presume that the baseball strike should cause a reduction in personal
income in MLB cities. If the typical team generates $300 million in economic activity (in 2000
dollars), the reduction in economic activity in each MLB city as a result of the strikes should be
roughly $100 million in each year. Adjusting for inflation, the loss of 34% of games in 1981
should result in income losses of $53.8 million per host city. The loss of 29.5% of games in
1994 should result in a loss of $76.2 million per host city, and the loss of 11.1% of games and
9
29.0% of attendance in 1995 should result in the loss of $77.0 million. Of course, in Chicago
and New York City, the losses should be double these figures since each city hosts two MLB
teams. As a percentage of the average host city’s income, i.e. personal income in the
metropolitan statistical area (MSA) in which the home stadium is located, the resulting losses
represent 0.151% in 1981, 0.099% in 1994, and 0.095% in 1995.
To attempt to measure the actual effect of the baseball strike, we have selected
explanatory variables from past models to help establish what income would have been in the
absence of the strikes and then compare these estimates to actual income levels to assess the
contribution of the team to the local economy. The success of this approach depends on our
ability to identify those variables that explain the majority of observed variation in growth in
income in those cities that host a MLB team.
One technique is to represent a statistic for a city for a particular year as a deviation from
the average value for that statistic for cohort cities for that year. Such a representation over time
will, in effect, “factor out” general urban trends and developments. For example, if we identify a
particular city’s growth in income as 10 percent over time, but cities in general are growing by 4
percent, then we would conclude that this city’s pattern deviates from the norm by 6 percent. It is
the 6 percent deviation that requires explanation and not the whole 10 percent for our purposes in
this study. Furthermore, if history tells us that a city that experiences a growth in income that is
5 percent below the national average both before and during a strike, then it would be misguided
to attribute that 5 percent deficit to the strike. If during the strike, the city continued to exhibit
income increases 5 percent below the national norm, the logical conclusion is that the residents
simply substituted other spending in lieu of baseball during the strike.
10
Given the number and variety of variables found in regional growth models and the
inconsistency of findings with regard to coefficient size and significance, criticisms of any single
model could logically focus on the problems posed by omitted variables. Any critic, of course,
can claim that a particular regression suffers from omitted-variable bias, it is far more
challenging to address the problems posed by not including key variables in the analysis.
In explaining regional or metropolitan growth patterns, at least some of the omitted
variable problem can be addressed through representing relevant variables as deviations from
city norms. This leaves the scholar with a more manageable task, namely that of identifying
those factors that explain city growth after accounting for the impact of those forces that
generally have affected national, regional or MSA growth. For example, a variable is not needed
to represent the implications of federal revenue sharing if such a change affected all cohort cities
in similar ways.
Following the same logic, other independent variables should also be normalized, that is
represented as a deviation from an average value for MSAs or as a fraction of the MSA average.
For example, a firm’s decision to locate a new factory in city i depends not on the absolute level
of wages in city i, but city i’s wage relative to those of all cities with whom it competes for labor
and other resources. What we propose, therefore, is an equation for explaining metropolitan
income growth which incorporates those variables that the literature identifies as important, but
specified in such a way that those factors common to MSAs are implicitly included.
Everything discussed in this section of the paper to this point is intended to define the
regression analysis that will be used to assess changes in income attributable to the 1981 and
1994 baseball strikes. Equation (1) represents the model used to predict changes in income for
11
host cities.
(1)
where for each time period t,MYt
i = % change in income in the ith metropolitan statistical area (MSA),nt = number of cities in the sample,Wt
i = nominal wages in the ith MSA as a percentage of the average for all cities in the sample,
Tti = state and local taxes in the ith MSA as a percentage of the average for all
cities in the sample,BOOMt
i = a dummy variable for oil boom and bust cycles for selected cities and years, TRt
i = annual trend,, = stochastic error.
For the purposes of our analysis the functional form is linear in all the variables included
in equation (1). This equation is calculated for each of the American MLB host cities. Toronto
and Montreal are excluded due to lack of available data. While the average income variable is
significant is each city’s regression equation, the remaining variables specified in equation (1)
are not necessarily statistically significant in each city’s regression equation. In these cases,
variables were removed until each remaining variable was significant at the 5% level. As is to
be expected with time-series analysis, auto-correlation was identified as a problem in the
regression models for most cities. Therefore, Cochrane-Orcutt regression was used in all cities
to eliminate the serial correlation.
As mentioned previously, rather than specifying all the variables that may explain
metropolitan growth, we attempted to simplify the task by including independent variables that
are common to cities in general and the ith MSA in particular. In effect we have devised a
12
structure that attempts to identify the extent to which the deviations from the growth path of
cities in general (E MYti /nt) and city i’s secular growth path (MYi
t-1) are attributable to deviations in
certain costs of production (wages and taxes), demand related factors (population, real per capita
personal income), and dummy variables for oil boom and bust periods as well as the region in
which the MSA is located. Equation (1) was used to predict the growth path for income, and this
predicted value was compared to the actual growth in income to formulate a conclusion with
regard to the effect the baseball strikes on income in MLB cities in 1981, 1994, and 1995. Of
course, the credibility of this procedure depends on a robust equation for predicting income
growth.
The Results of Model #1
We examined the economic impact of the baseball strikes using data over the period 1969
through 2000. The time period was chosen due to the availability of city by city income data.
Seventy-three cities constituted our sample, representing all MSAs that were on average the
seventy-three most populous in the country over this period and includes all MSAs that appeared
in the top sixty largest at any time during this period. The cities used are listed in Appendix 1
along with other information regarding the availability of data. The results of a regression for
Minneapolis-St.Paul using equation (1) are represented in Table 1. While each MLB city will
have different regression results, the Minnesota Twins (Minneapolis/St. Paul MSA) were used
for illustrative purposes.
A brief examination of the coefficients in Table 1 reveals some interest facts about
metropolitan area income growth in Minneapolis. As noted previously, not every variable is a
13
significant predictor of income growth in every city. For Minneapolis, neither taxes, wage, nor
lagged income growth is a statistically significant predictor of current income growth in the
model as specified. While wages tend to be a good predictor when conducting cross-city
comparisons, the rate of wage change within a city is generally too small to be a significant
predictor within a city over time.
The key statistic for our purposes is the difference between the actual growth in income
and that predicted for the city hosting the strike years. A complete listing of each city’s expected
income gains, realized income gains, and income gains above or below expected numbers is
shown in Tables 2a-2c.
As shown in Table 2, in no cities in any year did the strike emerge as a statistically
significant event at the 1% significance level. In thirty-six out of seventy cases, the increase in
income in the MLB city was lower than expected, while income gains were above the expected
amount in the remaining thirty-four cases. On average, the model predicted an increase in
income in host cities of 1.950% during strike years while the observed gains in income averaged
1.976%. The strike years produced an increase in income of roughly three-hundredths of a
percent above what would be expected, directly opposed to the ex ante estimates of decreases in
income of ten to fifteen-hundredths of a percent.
The magnitude of the variation of the estimates at first blush appear high. Some host
cities (New York City, 1995) exhibited billions of dollars in increased economic activity while
others (Detroit, 1981; L.A., N.Y.C., 1994) experienced billions of dollars in reduced economic
impact. The explanation for this range of estimates is simply that the models do not explain all
the variation in estimated income, and, therefore, not all the variation can be attributable to
14
baseball’s work stoppages. The standard error of the estimates from the models indicates
significant variation in the residuals from year to year for some cities. This heteroscedasticity
problem is particularly apparent in cities such as Detroit, Houston, and Miami where the
metropolitan economies are dominated by a cyclical industry: automobiles in Detroit, tourism in
Miami, and oil in Houston. For a large, diverse metropolitan economy, even a $100 million
dollar loss as a result of a MLB strike team is small portion of total annual economic activity for
the area. For example, in the New York City MSA, a $100 million event would represent only
0.04 percent of the city’s total GDP in 1995, a figure well below the standard error of the
estimate. While a $100 million event may not appear as statistically significant in any one host
city and is likely to be obscured by natural variations in the MSA’s economy, one should
expected that on average across the many MLB cities and numerous strike years, the average of
many $100 million losses will begin to appear as statistically significant. One can observe this by
adjusting the predicted income increases to assume a drop in MSA income of 34% of $300
million adjusted for inflation in 1981, 29.5% of $300 million in 1994, and 29.0% of $300
million in 1995 and standardizing the residual for each host city in each strike year. These
resulting standardized errors can be used estimate p-values. Essentially, the values in the
“Difference” column in Tables 2a-2c, adjusted for expected strike losses, are divided by the
standard deviation of the yearly residuals for the appropriate city. The mean of these
standardized residuals (= 0.13) is divided by the square root of 70 (the sample size) in order to
find a t-statistic with 69 (= n-1) degrees of freedom. The residuals in Tables 2a-2c (and hence the
mean of the standardized residuals) can be adjusted by assuming an economic impact larger or
smaller than the booster’s claims of $300 per team. The resulting p-values shown in Table 3
15
assume normality of the residuals. Using this methodology it is found that there is less than a 14
percent probability that the MLB teams provide $300 million or more in annual economic
benefits to the host community. In fact, the data suggest that the net impact of a MLB team is
approximately zero. Again it must be pointed out that due to the tiny magnitude of professional
sports in the scope of an entire metropolitan economy, a 90% confidence interval for the true
impact of a MLB franchise is quite large, ranging from a positive $449.6 million to negative
409.4 million.
Model #2
The key to the model presented in the previous section is to factor out experiences that
are common to all MSAs in order to be able to capture the effects of the baseball strikes. This
concept suggests another simple yet compelling method of calculating the impact of a MLB
sports franchise on a host city. One can simply calculate the ratio of total income in MLB cities
to the total income in a cohort of other metropolitan areas without MLB teams. While numerous
factors such as inflation, population, the economic business cycle, expectations about the future,
and seasonal variations can affect personal income in a city, most of these factors will affect
personal income in other metropolitan areas in a similar manner. Therefore, although it may be
difficult to predict income fluctuations in a city, if economic factors affect all cities in a given
sample in the same way, then the ratio of a particular city’s or group of cities’ incomes to the
incomes of cities in the sample as a whole should remain unchanged. If an event such as a strike
significantly decreases economic activity in the host cities, then the host cities’ incomes as a
percentage of incomes of the cities in the rest of the sample should decrease. By comparing the
16
MLB city/non-MLB city ratio in a strike period to other time periods, an decrease in income can
be inferred.
Several known variables will serve to shift the ratio and must be accounted for when
estimating the ratio. First, if the economies of MLB cities are growing at a faster rate than the
other cities in the sample, then the income ratio will grow over time. Thus, it is reasonable to
include a time trend variable in the model. This time trend variable can be inserted either as a
linear or a quadratic variable. The income ratio lagged any appropriate number of periods can
also be included if it is a good predictor of the current period ratio. Equation (2) represents the
model used to predict changes in the MLB/non MLB income ratio for our sample of
metropolitan statistical areas.
(2)
where for each time period t,Rt = ratio of income in MLB cities to all other sample MSAsTRt = annual trend,, = stochastic error.
The Results of Model #2
We examined the economic impact of the baseball strikes using data over the period 1969
through 2000. As before, the time period was chosen due to the availability of city by city
income data. Seventy-three cities constituted our sample, representing all MSAs that were on
average the seventy-three most populous in the country over this period and includes all MLB
host cities. To examine the 1981 strike, the total personal income of the 22 MSAs which hosted
the 24 American MLB teams was divided by the total personal income of the remaining 51
17
MSAs in the sample. Ordinary least-square regression was initially attempted, but was found to
have significant problems with auto-correlation as might be expected in a time-series analysis.
The Cochrane-Orcutt method was used to correct for this problem, and the results of a Cochrane-
Orcutt regression for equation (2) are represented in Table 4a.
An examination of the residual data for the regression model in Table 4a reveals that the
actual income ratio for the MLB cities to the non-MLB baseball cities was 0.28836% higher than
predicted by the model in 1981. As the personal income in the 51 non-MLB cities in the sample
totaled $1.319 trillion (in 2000 dollars) in 1981, a 0.28836% increase translates into a $3.80
billion gain to MLB cities during the strike year or $158.5 million for each of the 24 American
MLB teams in existence in 1981. Since the strike canceled 34% of the season, the $158.5 million
gain may be assumed to represent 34% of the value of the team leading to a total impact of a
MLB team on a host community of negative $466.1 million.
To examine the 1994 strike, the total personal income of the 24 MSAs which hosted the
26 American MLB teams was divided by the total personal income of the remaining 49 MSAs in
the sample. The results of a Cochrane-Orcutt regression for equation (2) are represented in
Table 4b.
An examination of the residual data for the regression model in Table 4b reveals that the
actual income ratio for the MLB cities to the non-MLB baseball cities was 0.5491% lower than
predicted by the model in 1994 and 0.1881% higher than predicted by the model in 1995. As the
personal income in the 49 non-MLB cities in the sample totaled $1.880 trillion (in 2000 dollars)
in 1994, a 0.5491% shortfall translates into a $10.323 billion loss to MLB cities during the strike
year or $397.0 million for each of the 26 American MLB teams in existence in 1994. Since the
18
strike canceled 29.5% of the season, the $397 million loss may be assumed to represent 29.5% of
the value of the team leading to a total impact of a MLB team on a host community of $1.346
billion. The personal income in the 49 non-MLB cities in the sample totaled $1.936 trillion (in
2000 dollars) in 1995, so a 0.1881% increase translates into a $3.642 billion gain to MLB cities
during the strike year or $140.1 million for each of the 26 American MLB teams in existence in
1995. Since the previous year’s strike reduced attendance by 29.0%, the $5.96 million gain may
be assumed to represent 29.0% of the value of the team leading to a total impact of a MLB team
on a host community of negative $483.0 million.
On average, the three models predict a total impact of a MLB team on host cities of
$132.3 million, or roughly one-half of the estimates provided by MLB boosters. Of course, the
wide range of estimates provided here, ranging from negative $483 million to a positive $1,346
million, serves to illustrate the difficulty of estimating the impact of a small business such as a
professional sports franchise on a typical large, diverse metropolitan economy.
Conclusions and Policy Implications
Major League Baseball teams have used the lure of economic riches as an incentive for
cities to construct new stadiums at considerable public expense. Estimates of the economic
impact of a MLB on host communities have typically been in the vicinity of $300 million. We in
general would urge caution with respect to these sorts of economic impact estimates, and our
analysis suggests that a figure of $300 is wildly optimistic. Our detailed city by city regression
analysis over the period 1969 to 2000 reveals that cities with MLB teams actually had higher
than expected income growth in the strike years of 1981, 1994 and 1995. While the range of
19
statistically likely net economic impacts for a MLB team on a host city ranges from a positive
$449.6 million to negative 409.4 million, a best guess at this impact is $16.2 million or roughly
5% of booster estimates. A second analysis of the ratio of income in MLB metropolitan areas to
non-MLB metropolitan areas over the period 1969-2000, again using the strike years of 1981,
1994, and 1995 as test cases, implied that the net economic impact of a MLB team on a host city
was $132.3 million. While this estimate is significantly larger than the estimate produced by the
first model, the figure is still less than half that suggested by most impact studies.
Cities would be wise to view with caution the economic impact estimates provided by
supporters of MLB. As a method of economic development, professional baseball, like Casey at
bat, strikes out. In addition, MLB cities should worry little about potential MLB work
stoppages. While baseball strikes do cause localized hardship, consumers find other outlets for
their spending leaving total city income relatively untouched.
20
APPENDIX
Table A1: Cities and years used to estimate model in Table 1 and 2
City Name 1969Population
1969Rank
2000Population
2000Rank
Wage Data availability Region
Akron, OH 676,214 59 695,781 77 1972-2000 Great LakesAlbany, NY 797,010 50 876,129 68 1969-2000 MideastAtlanta, GA 1,742,220 16 4,144,774 9 1972-2000 SoutheastAustin, TX 382,835 88 1,263,559 47 1972-2000 SouthwestBaltimore, MD 2,072,804 12 2,557,003 18 1972-2000 MideastBergen, NJ 1,354,671 26 1,374,345 44 1969-2000
(State data 1969-2000) Mideast
Birmingham, AL 718,286 54 922,820 67 1970-2000 (State data 1970-1971)
Southeast
Boston, MA 5,182,413 4 6,067,510 4 1972-2000 New EnglandBuffalo, NY 1,344,024 27 1,168,552 52 1969-2000
(Average of cities) Mideast
Charlotte, NC 819,691 49 1,508,050 42 1972-2000 SoutheastChicago, IL 7,041,834 2 8,289,936 3 1972-2000 Great LakesCincinnati, OH 1,431,316 21 1,649,228 34 1969-2000 Great LakesCleveland, OH 2,402,527 11 2,250,096 24 1969-2000 Great LakesColumbus, OH 1,104,257 33 1,544,794 41 1972-2000 Great LakesDallas, TX 1,576,589 18 3,541,099 10 1972-2000 SouthwestDayton, OH 963,574 42 950,177 65 1969-2000 Great LakesDenver, CO 1,089,416 34 2,120,775 25 1977-2000 Rocky MountainsDetroit, MI 4,476,558 6 4,444,693 7 1976-2000 Great LakesFort Lauderdale, FL 595,651 70 1,632,071 36 1969-2000
(State data 1988-2000) Southeast
Fort Worth, TX 766,903 51 1,713,122 30 1976-2000 (State data 1976-1983)
Southwest
Fresno, CA 449,383 79 925,883 66 1969-2000 (State data 1982-1987)
Far West
Grand Rapids, MI 753,936 52 1,091,986 59 1976-2000 Great LakesGreensboro, NC 829,797 48 1,255,125 48 1972-2000 SoutheastGreenville, SC 605,084 67 965,407 63 1969-2000
(State data 1969) Southeast
Hartford, CT 1,021,033 39 1,150,619 53 1969-2000 New EnglandHonolulu, HI 603,438 68 875,670 69 1972-2000 Far WestHouston, TX 1,872,148 15 4,199,526 8 1972-2000 SouthwestIndianapolis, IN 1,229,904 30 1,612,538 37 1989-2000 Great LakesJacksonville, FL 610,471 66 1,103,911 57 1972-2000
(State data 1988-2000) Southeast
Kansas City, MO 1,365,715 25 1,781,537 28 1972-2000 PlainsLas Vegas, NV 297,628 116 1,582,679 39 1972-2000 Far WestLos Angeles, CA 6,989,910 3 9,546,597 1 1969-2000
(State data 1982-1987) Far West
Louisville, KY 893,311 43 1,027,058 61 1972-2000 SoutheastMemphis, TN 848,113 45 1,138,484 54 1972-2000 SoutheastMiami, FL 1,249,884 29 2,265,208 23 1969-2000
(State data 1988-2000) Southeast
Middlesex, NJ 836,616 47 1,173,533 51 1969-2000 Mideast
21
(State data 1969-2000) Milwaukee, WI 1,395,326 23 1,501,615 43 1969-2000 Great LakesMinneapolis, MN 1,991,610 13 2,979,245 13 1972-2000 PlainsMonmouth, NJ 650,177 62 1,130,698 56 1969-2000
(State data 1969-2000) Mideast
Nashville, TN 689,753 57 1,235,818 49 1972-2000 SoutheastNassau, NY 2,516,514 9 2,759,245 16 1969-2000 MideastNew Haven, CT 1,527,930 19 1,708,336 31 1969-2000
(Average of cities) New England
New Orleans, LA 1,134,406 31 1,337,171 46 1972-2000 SoutheastNew York, NY 9,024,022 1 9,321,820 2 1969-2000 MideastNewark, NJ 1,988,239 14 2,035,127 26 1969-2000
(State data 1969-2000) Mideast
Norfolk, VA 1,076,672 36 1,574,204 40 1972-2000 (State data 1973-1996)
Southeast
Oakland, CA 1,606,461 17 2,402,553 21 1969-2000 (State data 1969-1987)
Far West
Oklahoma City, OK 691,473 56 1,085,282 60 1969-2000 SouthwestOrange County, CA 1,376,796 24 2,856,493 14 1969-2000
(State data 1982-1987) Far West
Orlando, FL 510,189 76 1,655,966 33 1972-2000 (State data 1988-2000)
Southeast
Philadelphia, PA 4,829,078 5 5,104,291 5 1972-2000 MideastPhoenix, AZ 1,013,400 40 3,276,392 12 1972-2000
(State data 1972-1987) Southwest
Pittsburgh, PA 2,683,385 8 2,356,275 22 1972-2000 MideastPortland, OR 1,064,099 37 1,924,591 27 1972-2000 Far WestProvidence, RI 839,909 46 964,594 64 1969-2000 New EnglandRaleigh-Durham, NC 526,723 73 1,195,922 50 1972-2000 SoutheastRichmond, VA 673,990 60 999,325 62 1972-2000 SoutheastRiverside, CA 1,122,165 32 3,280,236 11 1969-2000
(State data 1982-1987) Far West
Rochester, NY 1,005,722 41 1,098,314 58 1969-2000 MideastSacramento, CA 737,534 53 1,638,474 35 1969-2000
(State data 1982-1987) Far West
St. Louis, MO 2,412,381 10 2,606,023 17 1972-2000 PlainsSalt Lake City, UT 677,500 58 1,337,221 45 1972-2000 Rocky MountainsSan Antonio, TX 892,602 44 1,599,378 38 1972-2000 SouthwestSan Diego, CA 1,340,989 28 2,824,809 15 1969-2000
(State data 1982-1987) Far West
San Francisco, CA 1,482,030 20 1,731,716 29 1969-2000 (State data 1982-1987)
Far West
San Jose, CA 1,033,442 38 1,683,908 32 1972-2000 (State data 1982-1987)
Far West
Scranton, PA 650,418 61 623,543 84 1972-2000 (State data 1983-1984)
Mideast
Seattle, WA 1,430,592 22 2,418,121 19 1972-2000 (State data 1982-2000)
Far West
Syracuse, NY 708,325 55 731,969 73 1969-2000 MideastTampa, FL 1,082,821 35 2,403,934 20 1972-2000
(State data 1988-2000) Southeast
Tulsa, OK 519,537 74 804,774 71 1969-2000 SouthwestWashington, DC 3,150,087 7 4,948,213 6 1972-2000 Southeast
22
W. Palm Beach, FL 336,706 105 1,136,136 55 1969-2000 (State data 1988-2000)
Southeast
Complete data on population and employment was available for all cities from 1969 to
2000. This implies that data on employment growth and employment growth lagged one year
was available from 1971 to 2000. Data regarding state and local taxes as a percentage of state
GDP was available for all cities from 1970 to 2000, and was obtained from the Tax Foundation
in Washington, D.C. Wage data from the Bureau of Labor Statistics Current Employment
Statistics Survey was available for cities as described above. When city data was not available,
state wage data was used in its place. When possible, the state wage data was adjusted to reflect
differences between existing state wage data and existing city wage data. For MSAs that
included several primary cities, the wages of the cities were averaged together to create an MSA
wage as noted in Table A1.
The “Oil Bust” dummy variable was included for cities highly dependent on oil revenues
including Dallas, Denver, Fort Worth, Houston, New Orleans, Oklahoma City, and Tulsa. The
variable was set at a value of 1 for boom years, 1974-1976 and 1979-1981, and at -1 for the bust
years, 1985-1988. While this formulation does imply that each boom and bust is of an equal
magnitude, the variable does have significant explanatory value nonetheless.
Each city was placed in one of eight geographical regions as defined by the Department
of Commerce. The region to which each city was assigned is shown in Table A1. Employment,
income, and population data were obtained from the Regional Economic Information System at
the University of Virginia which derives its data from the Department of Commerce statistics.
23
REFERENCES
Baade, Robert A., 1996, Professional Sports as a Catalyst for Metropolitan Economic
Development, Journal of Urban Affairs, 18(1), 1-17.
Baade, Robert A. and Victor A. Matheson, 2000, An Assessment of the Economic Impact of the
American Football Championship, the Super Bowl, on Host Communities, Reflets et
Perspectives, 39(2-3), 35-46.
Baade, Robert A. and Victor A. Matheson, 2001 Home Run or Wild Pitch? Assessing the
Economic Impact of Major League Baseball’s All-Star Game. Journal of Sports
Economics, 2(4), 307-327.
Beaudry, Charles, 2002, Why Bother? The Economic Impact of a Professional Sports Franchise,
www.savetheexpos.com/impact.asp, accessed 9/5/2002.
Coates, Dennis and Brad Humphreys, 1999, The Growth Effects of Sports Franchises, Stadia,
and Arenas. Journal of Policy Analysis and Management, 14(4), 601-624.
Davidson, Larry, 1999, Choice of a Proper Methodology to Measure Quantitative and
Qualitative Effects of the Impact of Sport, The Economic Impact of Sports Events, ed.
Claude Jeanrenaud (Neuchatel, Switzerland: Centre International d’Etude du Sport), 9-
28.
Humphreys, Jeffrey, 1994, The Economic Impact of Hosting Super Bowl XXVIII on Georgia,
Georgia Business and Economic Conditions, May-June, 18-21.
Munsey and Suppes, 2000, Baseball Ballparks, www.ballparks.com/baseball, accessed
September 13, 2000.
Noll, Roger and Andrew Zimbalist, 1997, The Economic Impact of Sports Teams and Facilities.
24
Sports, Jobs and Taxes, eds. Roger Noll and Andrew Zimbalist, (Washington, D.C.:
Brookings Institution).
Oregon Baseball Campaign, 2002, Why MLB, Why Oregon, Why Now?
www.oregonbaseballcompaign.com/fact.htm, accessed September 5, 2002.
Porter, Philip, 1999, Mega-Sports Events as Municipal Investments: A Critique of Impact
Analysis, Sports Economics: Current Research, eds John Fizel, Elizabeth Gustafson, and
Larry Hadley (Westport, CT: Praeger Press).
Rosentraub, Mark, 1994, Sport and Downtown Development Strategy, Journal of Urban Affairs,
16(3), 228-239.
Saint Louis Regional Chamber and Growth Association, 2000, RCGA Estimates Economic
Impact of Cardinals Playoff Run, www.stlrcga.org/00_1002.html, accessed 9/5/2002.
Seigfried, John and Andrew Zimbalist, 2000, The Economics of Sports Facilities and Their
Communities, Journal of Economic Perspectives, 14(3), 95-114.
25
TABLE 1
Cochrane-Orcutt Regression Results for Income Data for Minneapolis MSA
Statistic/Valuea Coefficient Values and (t-statistics)b0 (constant) -1.3641 (-4.55)*b1 (MYt
i /ME MYti /nt) 0.9957 (15.06)*
b2 (MYti /MMYi
t-1) -b3 (MNt
i /MWti) -
b4 (MNti /MTt
i ) -b5 (MNt
i /MBOOM/BUSTti) -
b6 (MNti /MTRt
i) 0.0007 (4.56)*R2 .9019Adjusted R2 .8906F-statistic 119.37*
* Result was significant at the 99% level.
26
TABLE 2a
Actual vs. Predicted income growth in host cities, 1981
Year City Actual Growth Pred. Growth Difference t-stat Income Income gains 1981 ANA 3.513% 3.548% -0.035% -0.03 $ 56,438,972 $ -19,5781981 ATL 2.136% 3.241% -1.105% -0.97 $ 49,087,993 $ -542,2011981 BAL -0.115% 0.989% -1.104% -2.04 $ 50,832,822 $ -561,2851981 BOS 1.221% 1.613% -0.392% -0.34 $122,032,625 $ -478,4321981 CHC/WS -0.572% -0.262% -0.310% -0.40 $181,415,451 $ -563,0251981 CLE -1.281% -1.017% -0.264% -0.25 $ 53,244,147 $ -140,7451981 CIN -0.731% 0.206% -0.937% -1.12 $ 31,205,137 $ -292,4951981 DET -4.635% -1.884% -2.751% -1.54 $ 99,522,874 $ -2,738,0531981 HOU 8.445% 6.603% 1.842% 0.76 $ 79,054,855 $ 1,456,3121981 KC -0.909% 0.329% -1.238% -1.30 $ 32,721,681 $ -404,9711981 LA 1.351% 0.478% 0.873% 0.67 $192,451,311 $ 1,680,6071981 MIL -0.916% -0.452% -0.464% -0.55 $ 33,295,073 $ -154,6071981 MIN 0.467% 1.462% -0.995% -1.48 $ 54,596,942 $ -543,3611981 NYM/Y 1.636% -0.092% 1.728% 1.14 $208,286,617 $ 3,598,1961981 OAK 1.753% 2.190% -0.437% -0.39 $ 47,970,341 $ -209,5531981 PHL 0.278% -0.007% 0.285% 0.38 $108,739,253 $ 309,6651981 PIT -0.297% -0.827% 0.530% 0.63 $ 56,994,178 $ 302,1191981 STL 0.321% 0.056% 0.265% 0.45 $ 53,879,844 $ 142,5961981 SEA 1.635% 1.434% 0.201% 0.11 $ 44,690,981 $ 89,6271981 SD 3.440% 3.604% -0.164% -0.15 $ 45,080,844 $ -73,7201981 SF 2.814% 0.371% 2.443% 1.45 $ 50,621,535 $ 1,236,8051981 TEX 4.351% 4.483% -0.132% -0.10 $ 23,457,764 $ -30,964
Average(1981
1.087% 1.185% -0.098% -0.23 $ 76,164,602 $ 93,770
Average(1994)
2.035% 2.051% -0.016% 0.08 $ 100,596,670 $ -124,943
Average(1995)
2.733% 2.733% 0.184% 0.11 $ 103,375,273 $ 506,143
Average(Overall)
1.976% 1.950% 0.027% -0.01 $ 93,870,684 $ 160,167
27
TABLE 2b
Actual vs. Predicted income growth in host cities, 1994
Year City Actual Growth Pred. Growth Difference t-stat Income Income gains 1994 ANA 0.489% 0.070% 0.419% 0.36 $ 79,221,286 $ 331,6561994 ATL 5.367% 5.037% 0.330% 0.29 $ 96,041,711 $ 316,6121994 BAL 1.879% 1.089% 0.790% 1.46 $ 70,695,615 $ 558,1511994 BOS 2.556% 1.804% 0.752% 0.66 $180,893,292 $ 1,360,7791994 CHC/WS 2.619% 2.172% 0.447% 0.58 $241,648,864 $ 1,079,2801994 CLE 2.106% 0.988% 1.118% 1.06 $ 63,714,982 $ 712,4381994 CIN 2.241% 2.106% 0.135% 0.16 $ 43,137,519 $ 58,2301994 COL 3.939% 5.262% -1.323% -0.86 $ 54,842,926 $ -725,3941994 DET 4.376% 2.689% 1.687% 0.95 $128,602,419 $ 2,169,1791994 FLA 1.279% 0.202% 1.077% 0.34 $ 48,620,281 $ 523,4851994 HOU 2.192% 3.707% -1.515% -0.62 $102,378,735 $ -1,550,8761994 KC 3.272% 2.524% 0.748% 0.78 $ 46,084,284 $ 344,9341994 LA -0.893% 0.131% -1.024% -0.78 $240,785,132 $ -2,466,4511994 MIL 2.478% 1.913% 0.565% 0.67 $ 42,066,751 $ 237,7471994 MIN 4.072% 3.244% 0.828% 1.23 $ 83,882,610 $ 694,8971994 NYM/Y 0.414% 1.566% -1.152% -0.76 $293,618,843 $ -3,382,9291994 OAK 0.947% 2.256% -1.309% -1.18 $ 69,970,299 $ -916,2321994 PHL 0.355% 1.093% -0.738% -0.98 $147,523,768 $ -1,088,4891994 PIT 0.251% 0.997% -0.746% -0.89 $ 64,842,032 $ -483,9201994 STL 2.303% 1.768% 0.535% 0.91 $ 71,408,844 $ 381,9871994 SEA 2.459% 1.756% 0.703% 0.40 $ 70,806,403 $ 498,0081994 SD 0.443% 0.988% -0.545% -0.51 $ 69,532,351 $ -378,8201994 SF 1.182% 2.785% -1.603% -0.95 $ 66,499,827 $ -1,065,8991994 TEX 2.524% 3.076% -0.552% -0.43 $ 37,501,306 $ -207,007
Average(1981
1.087% 1.185% -0.098% -0.23 $ 76,164,602 $ 93,770
Average(1994)
2.035% 2.051% -0.016% 0.08 $ 100,596,670 $ -124,943
Average(1995)
2.733% 2.733% 0.184% 0.11 $ 103,375,273 $ 506,143
Average(Overall)
1.976% 1.950% 0.027% -0.01 $ 93,870,684 $ 160,167
28
TABLE 2c
Actual vs. Predicted income growth in host cities, 1995
Year City Actual Growth Pred. Growth Difference t-stat Income Income gains 1995 ANA 1.812% 1.631% 0.181% 0.16 $ 80,656,411 $ 145,6171995 ATL 5.519% 5.327% 0.192% 0.17 $101,342,437 $ 194,7711995 BAL 1.329% 1.314% 0.015% 0.03 $ 71,634,904 $ 10,4861995 BOS 2.457% 2.945% -0.488% -0.43 $185,337,596 $ -904,6981995 CHC/WS 3.647% 2.713% 0.934% 1.21 $250,460,769 $ 2,338,2371995 CLE 1.219% 1.517% -0.298% -0.28 $ 64,491,874 $ -191,9771995 CIN 1.464% 2.359% -0.895% -1.07 $ 43,769,041 $ -391,7441995 COL 5.568% 4.217% 1.351% 0.88 $ 57,896,381 $ 781,9701995 DET 2.707% 2.991% -0.284% -0.16 $132,083,299 $ -375,5141995 FLA 3.164% 2.711% 0.453% 0.14 $ 50,158,671 $ 227,2651995 HOU 4.753% 3.808% 0.945% 0.39 $107,245,066 $ 1,013,7481995 KC 2.529% 2.993% -0.464% -0.49 $ 47,249,955 $ -219,0351995 LA 1.247% 0.767% 0.480% 0.37 $243,788,401 $ 1,170,8711995 MIL 1.818% 2.171% -0.353% -0.42 $ 42,831,638 $ 151,0801995 MIN 3.615% 3.695% -0.080% -0.12 $ 86,915,313 $ -69,1731995 NYM/Y 3.402% 0.710% 2.692% 1.78 $303,606,418 $ 8,171,7011995 OAK 2.863% 2.907% -0.044% -0.04 $ 71,973,729 $ -31,4831995 PHL 1.456% 1.150% 0.306% 0.41 $149,671,812 $ 458,0951995 PIT 0.793% 0.890% -0.097% -0.12 $ 65,355,969 $ 63,6581995 STL 2.263% 1.753% 0.510% 0.87 $ 73,025,146 $ 372,7551995 SEA 2.657% 3.401% -0.744% -0.42 $ 72,687,993 $ -540,5281995 SD 1.292% 1.719% -0.427% -0.40 $ 70,430,368 $ -301,0831995 SF 4.772% 3.799% 0.973% 0.58 $ 69,672,963 $ 677,6711995 TEX 3.251% 3.705% -0.454% -0.35 $ 38,720,405 $ -175,791
Average(1981
1.087% 1.185% -0.098% -0.23 $ 76,164,602 $ 93,770
Average(1994)
2.035% 2.051% -0.016% 0.08 $ 100,596,670 $ -124,943
Average(1995)
2.733% 2.733% 0.184% 0.11 $ 103,375,273 $ 506,143
Average(Overall)
1.976% 1.950% 0.027% -0.01 $ 93,870,684 $ 160,167
29
TABLE 3
Probabilities for Various Levels of Economic Impact Induced by a MLB Team
Economic Impact Probability of such an impact or greaterhaving occurred
$441.6 million 5.00%$346.3 million 10.00%$300 million 13.49%$200 million 23.68%$100 million 37.16%
$0 52.50%negative 47.50%
30
TABLE 4a
Cochrane-Orcutt Regression Results for Income Ratio Data for 1981 strike
FINAL PARAMETERS:
Estimate of Autocorrelation CoefficientRho: -0.7029Standard Error of Rho: 0.14520
Cochrane-Orcutt Estimates
Multiple R 0.9995R-Squared 0.9990Adjusted R-Squared 0.9987Standard Error 0.0046Durbin-Watson 2.1512
Analysis of Variance:
DF Sum of Squares Mean Square
Regression 4 .47241 .11810Residuals 23 .00049 .00002
Variables in the Equation:
B SEB BETA T SIG T
Ratiot-1 1.39348 .10479 1.48841 13.2982 .00000Ratiot-2 -.74493 .09083 -.85725 -8.2014 .00000Time -.00663 .00127 -.73948 -5.2296 .00003Time2 .00010 .00002 .39059 5.1683 .00003Constant .51908 .09893 5.2472 .00003
31
TABLE 4b
Cochrane-Orcutt Regression Results for Income Ratio Data for 1994 strike
FINAL PARAMETERS:
Estimate of Autocorrelation CoefficientRho: -0.5833Standard Error of Rho: 0.16580
Cochrane-Orcutt Estimates
Multiple R 0.9995R-Squared 0.9990Adjusted R-Squared 0.9988Standard Error 0.0046Durbin-Watson 2.2484
Analysis of Variance:
DF Sum of Squares Mean Square
Regression 4 .47531 .11883Residuals 23 .00048 .00002
Variables in the Equation:
B SEB BETA T SIG T
Ratiot-1 1.39664 .11815 1.48051 11.8205 .00000Ratiot-2 -.73423 .11176 -.82991 -6.5697 .00000Time -.00671 .00160 -.69355 -4.2012 .00034Time2 .00010 .00002 .36184 4.3833 .00022Constant .54961 .13312 4.1019 .00044